Joint ensemble learning-based risk prediction of Alzheimer's disease among mild cognitive impairment patients

Description

This study used the UDS version 3.0 from the National Alzheimer’s Coordinating Center (NACC) (https://www.naccdata.org),This study used data from 27,418 patients from the UDS 3.0. We developed a risk prediction model for AD among MCI patients using NACC data. Data from MCI patients were obtained from the NACC. Importance ranking and the SHapley Additive exPlanations (SHAP) method for the Random Survival Forest (RSF) and Extreme Gradient Boosting (XGBoost) algorithms in ensemble learning were adopted to select the predictors, and hierarchical clustering analysis was used to mitigate multicollinearity. A total of 3674 subjects with MCI were included. Thirteen predictors were ultimately identified.This model can be easily understood by nonstatisticians. Furthermore, it can be used in primary medical institutions to screen high-risk populations for further examination and treatment, to reduce the incidence of AD, and to prolong the life of AD patients.

Steps to reproduce

In this study, we developed a risk prediction model for AD among MCI patients using NACC data. We have adopted a relatively novel screening approach for predictive factors. Importance ranking and the SHapley Additive exPlanations method for the Random Survival Forest (RSF) and Extreme Gradient Boosting (XGBoost) algorithms in ensemble learning were adopted to select the predictors, and hierarchical clustering analysis was used to mitigate multicollinearity.

Institutions

Categories

Funding

Licence